|
Jim,
Correct me if I'm wrong, but I don't see much
difference between![]() back- and forward testing.
E.g. if you test a daytrading system real-time for
5 days and then you forget the results and using exactly the same criteria,
backtest
those 5 days, the results should be the
same.
Systems stop working because markets change and
they have to be adjusted to new conditions. That's why backtesting often works
for the future and sometimes it
doesn't.
One could use testing period long enough to include
all kinds of markets, volatile, dead, etc. to get
a high probability of positive
system results over such period.
For my purpose, I pefer to adjust the system from
on market type to another as they come because that gives me much better
results.
Jan Philipp
__._,_.___
![]()
__,_._,___
|